Practically everyone who visits the Vatican stops to take a picture of the Swiss Guards. Ditto for the Queen's Guard at Windsor Castle, the Royal Life Guards at Rosenborg Castle in Copenhagen, and the Evzones at Greece's Tomb of the Unknown Soldier. Those multicolored uniforms may not have a place on the modern battlefield, where camouflage is far more important than panache, but they do attract the tourists.

If, by this measure, these ceremonial units have some value (albeit none militarily), why not have more of them? You could post the newly created Sartorial Guard at tourist locations that haven't been attracting enough foot traffic lately. And who knows, they might even attract more recruits into the real military. (Though I'm not sure what the career path is once you've held the rank of Feldweibel in the Swiss Guards.)

Obviously, I'm not being serious. Once you start manufacturing new ceremonial units, you cheapen the brand. You don't need more than one as a "loss leader" in the military, and there's no need to get the people who actually fight up in arms. Figuratively, that is.

Here's why managing a portfolio is critical for managing products. It wouldn't be hard to find some enterprising "champion" for a new Swiss Guards-ish unit who was willing to sew the uniforms and stand around looking fierce. (We call them re-enactors, and we don't put them on the public payroll.) No matter how much attention they attract, they'll still be a failure from a national perspective.

Companies are looking at their portfolios carefully. Arguably, they're re-thinking the concept of portfolio, with a greater emphasis on the value (measured along dimensions like business/IT alignment) the software assembly line is producing and less on how much it costs to run, how many workers should be on assembly line B today, or how fast the assembly line is running.

Vendors in the tools business have taken notice of this new demand. Some project portfolio management (PPM) vendors are trying to figure out how to reinvent themselves. Some ALM vendors are thinking about plugging into PPM. We may not know the exact contours of what ALM plus PPM would look like, but we're seeing the rough outlines already.

What's more clear is the responsibility of product management. For years, people in product management circles have been arguing over which noun best describes the job. Champion? Mini-CEO? Owner? Enabler? Quasi-MBA? They're all wrong, to the extent that they imply that the duty of the product manager is solely to the product. The PM needs to make business sense of development investments, not add further seats to the General Assembly.

That portfolio thinking has to come from somewhere. Among some of the best PM organizations I know, it's managed within the PM organization itself. The head of the PM team lays down the rules for how to justify, in business terms, any product idea. No clear market? Bzzzt, try again. Doesn't contribute to the overall value proposition? Bzzzt, I'm sorry. Being the champion, mini-CEO, lifelong companion, or BFF of a product is not always a good thing.

The point of setting these bars is not to squelch innovation or turn PM into the department of wet blankets. It's also not necessarily to judge the intrinsic worth of the idea. Instead, this set of questions is important for putting the new idea in the context of the larger portfolio. If you really want your product to succeed, demonstrate why customers will value it, why salespeople will sell it, and why partners will re-sell it. Or, in the IT world, why this idea is worth pursuing over something else.

If you're a product manager and you can't justify your product in the larger portfolio context, take a long look at the picture of the Swiss Guard in this post. Is this what you want? If, like the Swiss Guard, your product is a loss leader, go ahead and make your case on those grounds. If, like the Swiss Guard, your idea looks kinda cool but has no clear purpose at a portfolio level, don't be a champion, zealot, World's Biggest Fan, or Feldweibel for a product that shouldn't exist.